IIS team wins Best Paper Award
Lukas Cavigelli, Philippe Degen and Luca Benini have won the Best Paper Award at the 11th International Conference on Distributed Smart Cameras (ICDSC) for their paper "CBinfer: Change-Based Inference for Convolutional Neural Networks on Video Data" presented on 5 September 2017 in Stanford, CA, USA.
The authors propose a novel algorithm and implementation shown to be capable of providing a tremendous speed-up of 8.6x and a 10x higher energy-efficiency for convolutional neural network inference on static camera video data. This work paves the way for real-time video analysis on embedded smart cameras, enabling an on-site understanding of what is happening in the field of view, potentially alerting first responders.
Full text preprint: external pagehttps://arxiv.org/pdf/1704.04313.pdf
Congratulations!